|
From: | Daniel James |
Subject: | [Gnu-arch-users] Re: arch roadmap 1 (and "what's tom up to") |
Date: | Tue, 06 Jul 2004 21:13:08 +0100 |
User-agent: | Mozilla Thunderbird 0.5 (X11/20040306) |
Colin Walters wrote:
It seems that some scripting is required, so what you're suggesting will lead to other languages becoming a defacto part of the protocol.No, not at all. The project-specific hooks will add optional *convenience* functions. Like "tla submit". If you don't want to use the provided hook (e.g. echo blah | mail address@hidden), then you do it manually or write your own version for that project.
Okay. It seems to me that scripting should be more intrinsic that that, but I'll leave that point to people more involved enough in arch. I just wanted to discuss why, for me, an embedded language would be better than exec.
I also realise that if he was the kind of person who settles for what's available, arch would not exist.Yes, but arch had (and still has) very little in the way of competitors in the distributed revision control space. And what competitors exist are all really quite different, with major tradeoffs. Contrast to programming languages, where competition is extremely fierce, and reinvention of the wheel is common.
That's a very good point. Until recently, I han't really used arch in a distributed manner, which is why I didn't make that distinction. I was probably wrong to make that comparison.
In Furth's defence (in general, not specifically related to Arch), I do think it's doing something fairly unique, and I'm looking forward to seeing how it turns out.
thanks, Daniel
[Prev in Thread] | Current Thread | [Next in Thread] |